COSMIK-MPPI: Scaling Constrained Model Predictive Control to Collision Avoidance in Close-Proximity Dynamic Human Environments

arXiv cs.RO / 4/14/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper addresses the challenge of collision-free physical interaction between torque-controlled robot manipulators and humans in close-proximity, dynamic environments where safety constraints must be respected.
It proposes COSMIK-MPPI, a collision-avoidance framework that combines MPPI control with the RT-COSMIK human motion estimation system and a Constraints-as-Terminations method to enforce safety without relying on large penalty terms.
By treating constraint violations as terminal events, the approach aims to provide more reliable constraint satisfaction than additive-penalty schemes used in vanilla MPPI.
Experiments show COSMIK-MPPI reaches a 100% task success rate in real-manipulator tests with constant computation time of about 22 ms and substantially outperforms gradient-based MPC in the evaluated settings.
In simulations including infeasible scenarios, COSMIK-MPPI consistently produces collision-free trajectories, unlike vanilla MPPI, enabling robust human-robot shared workspace execution with an affordable markerless estimator.

Abstract

Ensuring safe physical interaction between torque-controlled manipulators and humans is essential for deploying robots in everyday environments. Model Predictive Control (MPC) has emerged as a suitable framework thanks to its capacity to handle hard constraints, provide strong guarantees and zero-shot adaptability through predictive reasoning. However, Gradient-Based MPC (GB-MPC) solvers have demonstrated limited performance for collision avoidance in complex environments. Sampling-based approaches such as Model Predictive Path Integral (MPPI) control offer an alternative via stochastic rollouts, but enforcing safety via additive penalties is inherently fragile, as it provides no formal constraint satisfaction guarantees. We propose a collision avoidance framework called COSMIK-MPPI combining MPPI with the toolbox for human motion estimation RT-COSMIK and the Constraints-as-Terminations transcription, which enforces safety by treating constraint violations as terminal events, without relying on large penalty terms or explicit human motion prediction. The proposed approach is evaluated against state-of-the-art GB-MPC and vanilla MPPI in simulation and on a real manipulator arm. Results show that COSMIK-MPPI achieves a 100% task success rate with a constant computation time (22 ms), largely outperforming GB-MPC. In simulated infeasible scenarios, COSMIK-MPPI consistently generates collision-free trajectories, contrary to vanilla MPPI. These properties enabled safe execution of complex real-world human-robot interaction tasks in shared workspaces using an affordable markerless human motion estimator, demonstrating a robust, compliant, and practical solution for predictive collision avoidance (cf. results showcased at https://exquisite-parfait-ffa925.netlify.app)